System Builder Marathon: Performance And Value Compared

Results: Adobe Creative Suite

Although the previous page appeared to favor Intel's architectures, its chart scaling was relatively linear. It seems like Adobe's After Effects might be similarly slanted. However, we have another explanation, based on previous experience. After Effects appears to penalize lots of cores and not enough memory, perhaps due to an allocation issue. By pairing six cores to 8 GB of RAM, performance really suffers. Meanwhile, my six-core machine with 16 GB of memory is afforded an opportunity to shine.

OpenCL-optimized filters appear to favor Nvidia’s GeForce GTX 770 over the GTX 760 found in both the $2550 and $650 machines. The program doesn’t appear to benefit from the expensive machine’s SLI array, but the combination of a lower-model GPU and a cheap CPU spell trouble for the $650 build.

Adobe Premiere rewards the $1300 PC’s newer architecture, but the $2550 PC’s extra cores still give it a slight advantage.

That doesn’t apply to Acrobat, which is single-threaded and consequently likes running on Haswell better than Sandy Bridge-E or Vishera.

Create a new thread in the US Reviews comments forum about this subject
This thread is closed for comments
42 comments
    Your comment
    Top Comments
  • itzsnypah
    It always seems that to be the best value in SBM you need the cheapest case, psu and motherboard and spend as much as you can on graphics.
    11
  • Other Comments
  • ryude
    Value is one thing, but when it comes to gaming you have to build for a minimum acceptable framerate and graphical fidelity. Once you factor those in you see that the $1300 build is indeed the value leader. The $650 build cannot play all games at 1080p60, high settings, and decent AA.
    -3
  • itzsnypah
    It always seems that to be the best value in SBM you need the cheapest case, psu and motherboard and spend as much as you can on graphics.
    11
  • rolli59
    Agree with the value pick and now looking to see what you guys did for $350!
    3
  • Martell1977
    So at what price point does diminishing returns really kick in, approximately? Would spending a little more on the GPU for the $650 still be a solid value add?

    Basically, at what point between $650 and $1300 does the price/performance ratio seriously diminish?
    3
  • Crashman
    Anonymous said:
    It always seems that to be the best value in SBM you need the cheapest case, psu and motherboard and spend as much as you can on graphics.
    That doesn't stop me from setting a minimum quality standard for the high-end build, that comes at a higher price than the minimum performance standard. And, it doesn't stop me from adding a secondary storage drive, because these are things that the owner of this system would expect to have. I go into this knowing that I'm "wasting" money on quality, features and convenience items, and it doesn't bother me at all :)
    Anonymous said:
    So at what price point does diminishing returns really kick in, approximately? Would spending a little more on the GPU for the $650 still be a solid value add?

    Basically, at what point between $650 and $1300 does the price/performance ratio seriously diminish?
    One of our SBM's focused on that question. It's currently somewhere around $700.
    9
  • ingtar33
    Anonymous said:
    Value is one thing, but when it comes to gaming you have to build for a minimum acceptable framerate and graphical fidelity. Once you factor those in you see that the $1300 build is indeed the value leader. The $650 build cannot play all games at 1080p60, high settings, and decent AA.


    you didn't bother reading the benching at all apparently. The 650 build was way over 60fps in all titles on ultra settings at 1080p except for far cry (it was even over 60fps on skyrim, which really hates amd cpus). Far Cry 3 has always been a gpu melter in the category of crysis 3; so it shouldn't be surprising a 760gtx can't max fc3 on ultra at 1080p. It doesn't in any other bench anywhere either. And fc3 was clearly playable on ultra at 1080p (30-40fps). Personally if i built a $650 machine and it killed every game i threw at it at 1080p and 60fps i'd call it a day. there really isn't a reason to spend more on your hardware unless you're going to spend a fortune on better/multiple monitors with bigger resolutions...

    Computer tech has come a long way, that we basically have a mainstream gaming platform at 1080p for $650 is a great thing.
    8
  • nerrawg
    Anonymous said:
    Value is one thing, but when it comes to gaming you have to build for a minimum acceptable framerate and graphical fidelity. Once you factor those in you see that the $1300 build is indeed the value leader. The $650 build cannot play all games at 1080p60, high settings, and decent AA.


    Actually what you realise is that the CPU on the $650 build is probably good enough for a GTX 780/AMD7970 or 2 GTX 760's in SLI. With that added expenditure of only $150-300 you could play anything you want to at 1080p without the PC breaking a sweat. It goes to show that, while the AMD Piledrivers are far behind intel's quad core K series, they can still represent decent value for a gaming PC. Nice article.
    6
  • m32
    ingtar33, I agree with you. In the coming months your $650 is going to get you more with AMD's 7000 series price dropping. More money is great if you got it. Get your Titans if you can! Average folks are going to be sticking to the sub $850 range.
    0
  • icerider
    Great SBM guys. Would have preferred to see the $650 machine get this,
    GIGABYTE GV-R795WF3-3GD Radeon HD 7950 3GB for $224 ($199 after rebate)
    and spent the additional money on a ,
    COOLER MASTER Hyper 212 Plus
    which gives plenty of headroom to take a fx-6300 to an easy 4.5 ghz OC with low temps. Just built my first 2 FX-6300s this way with absolutely no problem.
    With an extra Gig of graphics memory, comparable gpu oc ability and framerates and a solid OC on the cpu I think this system would be an easy walk away winner.
    0
  • bemused_fred
    OK, OK, OK. Could someone please explain why there are benchmarks in this at all? Does anyone ever expect the $650 PC to come near the $1300 and $2550 ones withs superior hardware? It's a complete waste of time to make those graphs, because a 5-year-old can probably tell you what the results will be: the more expensive stuff gets you better performance.
    -8
  • Crashman
    Anonymous said:
    OK, OK, OK. Could someone please explain why there are benchmarks in this at all? Does anyone ever expect the $650 PC to come near the $1300 and $2550 ones withs superior hardware? It's a complete waste of time to make those graphs, because a 5-year-old can probably tell you what the results will be: the more expensive stuff gets you better performance.
    1.) You need gaming benchmarks to PROVE that the slowest system can play these games at high setting.
    2.) You need application benchmarks to SHOW performance proportionality.

    You need those things. We don't. We could just discuss the results without showing them, but that discussion wouldn't make sense to you. And, you might be the first one to ask what we're hiding by not showing the charts.
    7
  • Mitche01
    This article raises an interesting point about workstations. I would like to see a Workstation System Builders Marathon, looking at budget, midrange and high end workstation systems.
    3
  • Rob Z
    Guys... another great build off. It does prove that you DONT need to have the most expensives parts on the market to put together a very capable gaming rig. When you look at your trying to set a specific goal of 1080p at 60fps you can really get creative on what parts you can use to accomplish the task. It may be great to say I have an i7 processor with tri- sli 760 cards but fo the $$$ you invested in it unless your trying to reach some off the cart resolution it showed that its not worth the money compared to a single card or even a dual ski setup. would have loved to see a better cooler or sealed water cooler on the fx-6300 and how much more performance you could have tweaked out of it (knowing that it wouldn't catch up to the i5 and definitely not the i7)
    2
  • vertexx
    Paying a premium for performance is nothing new. Take automobiles, for example:

    2013 Ford Taurus SHO
    MSRP: $28,900
    0-60 time: 5.1
    Relative Price: 100%
    Relative Perf: 100%
    "Value": 100%

    2013 Ford Shelby GT500
    MSRP: $59,200
    0-60 time: 3.6
    Relative Price: 205%
    Relative Perf: 142%
    "Value": 69%

    2014 Porsche 911 Turbo S
    MSRP: $97,350
    0-60 time: 2.9
    Relative Price: 337%
    Relative Perf: 176%
    "Value": 52%

    I suppose that means that true "value" is more than just what the numbers say. Price/performance is one thing. "Value" is something much more complex in the mind of the buyer.
    1
  • midnightgun
    This article pretty much sums up the state of gaming right now and computer sales in general. For one monitor gaming by the numbers, there really isn't a reason to spend more then a grand or 1500 on a machine. The 650 machine did most just fine.

    I have a feeling this is mostly due to the consoles pretty much dictating visual progress in most cases. As a result there isn't a need to increase performance of machines much, so Intel improves efficiency for other markets.

    As a result, people don't really get new machines as often cause there isn't much of a need. Hell I'm sitting on a Core i5 750 and only just recently started considering a new machine. I'll build a new one next year... When my machine crosses the 5 year old mark...

    And Tech companies sit and wonder why pc sales have slowed down? :\
    2
  • bemused_fred
    Anonymous said:
    Anonymous said:
    OK, OK, OK. Could someone please explain why there are benchmarks in this at all? Does anyone ever expect the $650 PC to come near the $1300 and $2550 ones withs superior hardware? It's a complete waste of time to make those graphs, because a 5-year-old can probably tell you what the results will be: the more expensive stuff gets you better performance.
    1.) You need gaming benchmarks to PROVE that the slowest system can play these games at high setting.
    2.) You need application benchmarks to SHOW performance proportionality.

    You need those things. We don't. We could just discuss the results without showing them, but that discussion wouldn't make sense to you. And, you might be the first one to ask what we're hiding by not showing the charts.



    Gee, if only there had been extensive benchmarks on all three systems for the last three days. Nah, that's crazy talk.

    Also, I really don't need benchmarks to know about performance proportionality. I've been into computer hardware for more than 15 minutes, so it's not exactly news to me that the entire market is subject to diminishing returns. I can't even remember the last time when a $400 CPU or GPU offered more than double the performance of a $200 one.
    -8
  • ojas
    Nice SBM. Next few quarters could get even more interesting. Keep it up! :)


    p.s. Please have newer games next year! 2013 titles should be a min requirement.
    2
  • the1kingbob
    Good article, as always. I have never quite understood why the $650 machine is expected/compared as such high resolutions. Ultra resolutions come from monitors costing quite a bit of money, but I guess you could cheap out on a machine for a fantastic monitor...
    2
  • Cache
    I'd be curious to know the expected lifespan/use of these systems over time. The budget build works well for now, certainly, but there are always more demanding game the following year and I'm not sure budget future-proofs for more than two years if gaming is set as a priority (at higher resolutions).
    1
  • cmi86
    Kind of lame we didn't see the full productivity suite used in this article which would have even further exclaimed the 6300's value vs the i5 in higher threaded workloads where it's not uncommon for the 6300 to out perform the i5. The results are already there you just have to cross reference them from the 2 different articles as opposed to having it all in this article for a final review.
    2